On Friday, February 1, 2002, at 06:45 AM, Gareth Coltman wrote: > This is going to be a common problem as far as I can see. In our case we > want to return a PDF (actually an FDF with an internal PDF hyperlink) > as the > result of a form fill. So I need to use the action to parse the form > data > and generate the FDF. When I actually write it onto the output stream is > obviously the point here.
I would want to separate this into two components. The action should capture the user input. And a subclass of RawScreen or the T3 equivalent should be responsible for the PDF generation. Perhaps the action captures input and puts it into a context that gets used by the PDFScreen, or something like that. In other words, treat the PDF generation as analogous to one of the template mechanisms. -Eric -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
